Where our models compound.
Each vertical gets its own motion. The keywords, the intent, the assistant prompts, the cadence — engineered for the category, not the average.
Fintech compounds on trust keywords. Gaming compounds on a 96-hour launch window. Media compounds on the second-session prompt. Same store, different physics.
ChatGPT, Gemini and Perplexity have distinct citation habits per category — reviews for fintech, Reddit for gaming, editorial press for media. The evidence base we build reflects that.
Each vertical maps to one of seven engagement models (M-01 → M-07). You know which one you're on, which cadence it runs, and what ships each week.
Fintech apps live and die on trust keywords.
Category leaders don't just rank for "budget app." They own the "without fees," "for freelancers," "in India" long tail — and the assistant answer when users ask which one to trust.
Health apps get recommended, not just searched.
Users ask assistants what to install for anxiety, sleep, macros, and offline meditation. If your metadata doesn't teach the machine what your app is for, someone else's does.
E-commerce apps compound the long tail — or they cap out.
The head keyword ceiling is real. Winners rewrite the description body every quarter to widen the semantic surface, and treat CPPs as a full-funnel asset — not screenshot A/B.
Edtech installs happen when the exam is next week.
Search intent is seasonal and locale-specific. Localised keyword panels, per-country screenshot narratives, and native-reviewer sign-off — not translation.
Gaming installs live inside four days of a launch window.
Featured slots, chart velocity, review sentiment in the first 72 hours — the compounding is different. Everything hinges on the pre-launch keyword panel, the icon that survives a thumbnail, and a review-response voice that doesn't sound like a support script.
Media apps live or die on the second-session prompt.
Users install for one show, one podcast, one article — then decide within 24 hours whether the app earned a place on the home screen. Store CVR is table stakes; the win is metadata + onboarding that pre-frames the second session.
The motion is the same. The vocabulary is what changes.
Productivity, dating, travel, developer tools — we've shipped in every category adjacent to these six. The free audit tells you which of the seven models fits before you commit.